Characterizing consumption patterns by food frequency methods: core foods and variety of foods in diets of older Americans.

This study was designed to explore the applications and limitations of two relatively simple and inexpensive scoring systems currently used to describe food patterns. The approaches examined were variety index and core foods. The sample, stratified by age and sex, consisted of adults more than 54 years old (no. = 4,983) who were interviewed in the 1977-1978 Nationwide Food Consumption Survey. Variety index was determined by counting the number of different foods consumed over 3 days. Core foods were identified by tallies of the most frequently mentioned foods and by calculations of frequency use scores. The variety of foods consumed was smallest for the oldest groups. Core food lists did not vary markedly with age, sex, or analysis by tally or score. The items consumed on at least 1 of the 3 days by approximately 50% of the persons surveyed were: whole milk, ground coffee, instant coffee, white bread, potatoes prepared by methods other than frying, margarine, and sugar. Although the variety index approach provides a general sense of nutrient diversity of a diet, core food lists with weighted scores appear to be more applicable when the relationship between food usage and food-related behaviors is studied and changes in eating patterns are monitored.